Urbanization in Ethiopia: Study on Growth, Patterns, Functions and Alternative Policy Strategy Tsegaye Tegenu,
Department of Human Geography,
Stockholm University
Stockholm, 2010 1. Background and Purpose
Forty years ago Gamst characterize Ethiopian civilization as “Peasantries and Elites without Urbanism”. He was referring to a presence of a social organization and high culture civilization in which cities were absent (Gamst 1970). Thirty years later, researchers were still talking about “urban deserts” in Ethiopia, despite the acceleration of urbanization, with an average growth rate between 4,4 and 5,6 per cent a year (Golini, 2001). A recent urbanization study that used an agglomeration index as a methodology concluded that “a large share of the population still resides more than 10 hours travel time from an urban center” (Schmidt and Kedir, 2009). Even though the background to the state of urban growth and urbanization is related to the predatory nature of the Ethiopian state (Tegenu 1996), in this paper the focus is on a period of globalization and start of demographic transition in Ethiopia (Malmberg 2008, Tegenu 2004).
The main purpose of this study is to examine the causes, patterns, consequences and policy implications of the ongoing urbanization in Ethiopia.
• First, it provides statistics about the scale and nature of the urban population change and the demographic causes that contribute to this change. To understand the speed of change comparison is made with other African countries;
• Second, it attempts to explain the level, pattern and trends of urbanization by examining the links between the function of towns and government policy;
• Finally, it provides policy suggestion by highlighting the consequences of rapid urbanization, reviewing of government policy and considering causes for the low level of urbanization.
After briefly presenting the basic framework of interpretation in section 2 below, the study discusses the time setting and driving forces behind rapid urbanization in Ethiopia in sections 3 and 4, the economic aspect and characteristics of urbanization in section 5, effects of rapid urbanization on employment, public service provision and spatial organization of towns in section 5. Sections 6 and 7 offer final conclusion in the form of development strategy and policy implication.
1
2. Determinants and Nature of Urbanization
2.1. Definition and Approach to Causes of Urbanization
A population that is urban is one in which vast numbers of people are clustered together in very small areas called towns and cities. “Urbanization is a process of population concentration. It proceeds in two ways: the multiplication of points of concentration and the increase in size of points of concentration.” (Oluwasola 2007). The physical growth of urban areas can be explained demographically and functionally. While demographic definition of urbanization is restricted to factors such as population size and density, the economic functional definition refers to the territorial concentration of productive activities (industries and service) rather than population. In this study urbanization is examined both demographically and functionally.
There are, however, three overlapping questions of considerations in explaining the synthesis between an increase in the urban share of total population and a structural economic change in a given country. The first set of questions refers to the identification of the demographic and economic drivers of change and how those drivers affect the number and size of urban centers? The second question relates to the identification of enabling factors which determine where the urban centers can be clustered (i.e., spatial distribution of urban centers). While the first question (demographic pressure) affects the increase in the overall growth rate and size of the urban population, the second question (institutional/policy responses) affect mainly the function of the towns not their demographic size. The third question relates to problem of periodization. Urbanization is not unique; it has been there all the time. Is there a time setting which gives particular characteristics to the urbanization process under consideration?
My models of urbanization try to associate population and economic changes in a time setting that is divergent from the past. The historical period in which the changes have occurred is known by the name demographic transition, which refers to the process through which countries’ birth and death rates pass from high levels to low levels. The decline of mortality usually precedes the decline in fertility, thus resulting in rapid population growth during the first stage of the transition. My point of departure is that during the stage of rapid rural population growth (high births than death) the determinant of urban growth is net rural urban migration than natural increase of the urban population. Rural net outmigration follows a pattern which at first increases, then passes through a maximum, and decreases thereafter
2
(see “mobility revolution” hypothesis in Ledent, 1982 and Figure 1). Increase in the number of small towns and size of urban population as a result of net internal migration is called here migration-led urbanization. Later on when rural fertility rates drop and when the main sources of urban growth is accounted by natural increase, urbanization (increase in the urban share of total population ) is the function of economic growth (growth-driven urbanization)
Source: Adapted from the figures given by Ledent 1982
Migration-led urbanization is rapid. It increases both in number of towns and proportion of the population located in the urban areas and this is mainly a function of the demographic transition. Its’ growth rate is faster than the financial and human resources available for managing its development. The population is increasing without an increase in industrial production and rising incomes. In the industrialized countries, urbanization took many decades creating conditions for the proliferation of tradable manufacturing and services and resources for facilitating and managing the provision of services to residents. Urbanization that has the capacity to improve the quality of life through state-market responses is growth driven.
2.2. Nature of Migration-led Urbanization
In the literature migration is discussed in the paradigm of push-pull factors. Push factors refer to the specific forces in individuals lives that lead to the decision to migrate outside of its
3
normal place, while pull factors refer to those that lead an individual to select one destination over another once the decision to migrate has been made. This type of analysis provides empirical fit to understand the circumstances that trigger the motivation for migration. But in studying urbanization my interest is not on the motivational framework (desire and choice) at the individual level. My interest is at the macro and local levels at which the causes (push/pull factors), levels, streams, intensities and forms of migration affect the number and size of points of population concentration.
Migration -led urbanization is characterized by a) an increase in the number of towns (multiplication of the points of concentration); b) gradual market integration of the different components of the agriculture economy (production, land and labour) and development of the rural market functions (handicraft, trade and service); and c) unbalanced growth in size and spatial distribution of towns (“distorted location incentives”).
There are different sources for an increase in the number of towns. Towns can grow from settlements physically in existence for many years (regional and zonal administrative centers and farm settlements). This type of urban growth is mainly related to nature of the land resource reallocation that increases rural-rural mobility/migration. In rural areas land could be acquired through inheritance, parental sharing, redistribution, forest clearing, renting and share cropping. These forms of output (increase through area extensification) increases size of some rural settlements which at certain threshold can be labeled as towns.
New towns can also grow as a result of the surplus labor mobility and occupational changes. The surplus labor consists of rural labor engaged in “elementary occupation”, the underemployed of the self employed farm labour (subsistence labour) and the new young adult who enter the labour force for the first time. As a result of land scarcity (due to fragmentation and entry barriers to share cropping and land rent) surplus rural labor migrates to urban centers. 1 This is often accompanied by occupational changes from farm to non-farm activities (petty trade, handicraft, transporting, mining, selling of wood, local brewery, etc.).
Without being forced to be mobile, subsistence households may also be forced to undergoing occupational changes from farm to non-farm activities as the total household
1 Higher population density with fixed technological and/or poor agro-ecological factors results in diminishing returns to labour. Increasing application of labour on fixed amount of land eventually leads to decreasing levels of output per unit of input. Under such conditions household either intensify agriculture or migrate. 4
consumption requirements increases due to their reproductive behavior. Under such circumstances households depend on the village market or nearby settlement areas for the sale of their labor and/or products. The intensity of non-farm activities by the subsistence households leads to an establishment of small market towns.
New towns can grow as a result of planned settlement activities of governments. Government rural settlement program is another factor which can contribute to the growth in the number of small towns (Silberfein, 1998). To save drought victims the government might move farmers out of the densely settled areas and settle them in another area. Villagization programs for the purpose of extending such services and facilities as schools, clinics and water supplies to the rural people may lead to an increase in the number of towns.
It is apparent that migration affects the overall demographic size of the towns. But its effect on the point of concentration (on the structure and spatial distribution of towns) is mediated through other factors. There are two enabling factors. The first is related to the nature of the government rural-urban biased policies (institutional/state responses). Development in transport, expansion of commerce and small manufacturing industries, re- organization of administrative set ups and decentralization affects the structure and spatial distribution of urban centers.
Some states take urban bias policy while others take rural bias policy. This depends on the historical nature of the state, the socio-cultural origin of the power elite, its ideology and political interest. 2 Following the colonial period, most African states followed urban bias policy since independence. 3 The urban bias policy is a result of the state’s effort to reverse the causality that retards urban growth and the reasons can be administrative, fiscal interest (investment) and/or creation of supply point to international trade (this is often discussed in terms of interdependent urbanization). 4
2 There are different ways of government’s stimulation of urban growth: high expenditure on town infrastructure and services at the expense of rural areas, supporting minimum wage rates and employment protection, subsidizing industries, food and housing to registered urban dwellers. 3 Urban bias can be a result of administrative, commercial and/or industrial interest. During colonial period cities were established as centers of administration, commerce, mining (gold, diamond and manganese), and extraction of export raw materials (cocoa, timber, cotton, etc.) from rural areas . 4 Some post independent African states have pursued the city bias policy without trying to restructure the impacts of the colonial space-economy. In the post colonial period the export of raw materials was the source of 5
Urban-bias policy has an effect on internal migration in the sense that better opportunities created by the urban policy --infrastructure development to extract agricultural production and the artificially low food price in urban area-- encourages rural migrants to towns. But urban- bias policy by itself does not affect the overall demographic size of the towns. In the absence of changes in the subsistence economy and growth in the rural labor force, government urban policy (pull factor) does not by itself lead to an increase in the overall growth rate and size of the urban population. The policy is not such that it attracts large out migrants to the extent of over-urbanization. Urban-bias policy affects the balance in the distribution of the urban population. If the policy is correct it leads to balanced distribution and equitable growth among towns. It does not lead to increase in the overall growth rate and size of the urban population.
The factors of absolute advantage are the second set of enabling factors which affect the distribution and cluster of urban centres. These factors are determined by agro-ecological settings (such as rainfall amount, soil quality, topography and altitude). These factors are physical given affecting the nature of crops and cropping patterns. They affect what farmers can or should produce. In a higher altitude, for instance, farmers cannot produce teff crop, since the climate is suitable only for growing crops such as barely and wheat. In highland areas people can generally pursue an agricultural pattern of livelihood, while in the lowlands they can grow few crops and will be either pastoralists or agro-pastoralists, and so on.
In a given agro-ecological factors households share similar climatic, topography, altitude and soil properties. Under such conditions and given similar demographic situation households can be similar with respect to the crops they are growing and the choice they have in labour allocation. In regions where households make the choice for cash crop and/or selling of their labour, migration is mainly intra-regional. Local movers are predominant than far distant migrants and this has positive effect on the size development of towns and their consequent rankings.
2.3. Nature of Growth-driven Urbanization
Growth-driven urbanization is a result of the productivity of manufacturing firms and an increase in tradable services. Manufacturing firms and tradable services are much less land the state economy and some of the modifications did not alter the policy importance given to city and town development at the expense of the rural areas . 6
intensive than agriculture (Henderson 1974) and they have the potential of spatial concentration for various reasons including lower transportation costs, access to natural resource inputs and increasing returns (WDR 2009). Both sectors can absorb the surplus labour in rural areas. Manufacturing productivity not only increases employment growth, but also stimulates the agricultural productivity through supply of inputs and relief of labour burden. Tradable services are distribution services that have lower transportation costs and relatively high consumers of the service and this sector can absorb growing rural labor force as the trade increases (Deng, et al. 2008).
The concentration of production and consumption in a given area, however, depends on institutional changes which facilitate factor mobility. It is important to identify the role of the state and the market in finding a policy balance and link between the rural and urban development. The state should be able to create the following enabling environments: it should desist from printing money and over spending to ensure macroeconomic stability; create a banking capacity appropriate for investment; provide infrastructure services, good governance and efficient judiciary system. These institutional changes create the necessary opportunity conditions for market forces (including decentralized local governments, private and corporate units (merchant and craft guilds in European proto industrialization and village and township enterprise as in the case of rural industrialization in China). Given favorable institutional environment local and international capital investment, labor force growth and technical progress are some of the contributing factors to the level of geographical concentration and spatial economic development. International financial and trade relations and situation can positively or negatively affect the urban-bias policy. 5
The democratic network governance created between the state (power elites) and the market forces can lead to an interest to balance urban and rural interests. The state for its legitimacy purpose (providing millions of jobs for urban people) and the market for its
5 Capitalism produces urbanization by concentrating production and consumption in locations that afford the greatest economies of scale, agglomeration and linkage, and where control over sources and supply can be exercised with maximum effectiveness, at least cost (Johnston, 1980). International economic relation facilitates the concentration of population around points of supply for an exchange of industrial and agricultural production. Cities, especially national capitals and those with major ports or international airports, offer overwhelming advantages for profitable investment, affording wide access to cheap labor and to domestic markets . 7
expansion can work both for the urban and rural interests. In some instances the state can stand above the sectional interest of both urban and rural population. This happens in a situation where there is a developmental state that strives for the objectives of economic efficiency and effectiveness. This type of state considers the wider interest of the society as a whole such as accumulation of capital for investment in heavy industries and infrastructure development. In that case both the rural and urban areas pay the costs of investment and growth. 6 Developmental state can act to further its own interest as a distinct social force. To stay in power, it has to work both for the market and the society and for the interest of the rural and urban population. Developmental states, which focus on comprehensive structural transformation of the economy and raison d’état, can transcend the rural-urban differences and cleavage between urban and rural bias policy.
3. Overlapping of Demographic Transition and Urban Transition
3.1. The Context of Urbanization: Population Growth and Pressure Distribution In Ethiopia the demographic transition (with mortality rates falling first, and fertility rates falling subsequently after a period of very rapid population growth) and the urban transition (with populations becoming increasingly urban) have coincided in time. The demographic transition has started since the 1950s, while rapid urbanization began since the 1960s. When the rapid phases of the two transitions overlaps a high urban growth rates is experienced as we shall see later.
Ethiopia is now on the onset of a demographic transition. Since the beginning of the 20th century, Ethiopia is experiencing a high natural population growth rate compared to the previous long periods where there was virtually no long-term population growth due to very high and varied birth and death rates. Population had grown more than five times since 1900, three times since 1955 and had doubled since the early 1970s. By the year 2008, the total population of the country was 79 million, second most populous country in sub-Saharan Africa after Nigeria (CSA 2007). Each year an estimated 2 million persons are added to the population. According to the projection made by CSA, the population increases at a rate
6 The role of the state should go beyond the creation of necessary institutional opportunities. The state must invest in strategic industries (such as chemical and machine industries) which can provide hardware and inputs to small scale enterprises.
8
reaching a minimum of 2,3% during the 2015-2020 period7 (NOP 2000). If the projection is borne out, the population of the country will increase from its present level of 79 million persons to 137 million persons by the year 2030 (Figure 2).
Source: Population Division, DESA, United Nations
Population is quite unevenly distributed in the various regions of the country. 81% of the total population is concentrated over 50% of the total landmass. Population density at the regional level varies from less than 10 persons per sq. km. to 4572 persons per sq. km. in Addis Ababa, the capital city. Nearly 80% of the population lives on only 37% of the total area of the country, while the remaining 20% lives on 63% of the country’s land area (Teller et al. 2007). The Ethiopian population has traditionally been highly concentrated in the highlands. The lowlands are very sparsely populated mainly because of malaria and other vector borne diseases.
This population dot map in Figure 3 presents total estimated population for 2004 based on the 1994 Census. Note that total population includes both agricultural and nonagricultural populations. Each dot represents 500 persons. The clustering of populations in the highlands
7 World growth rate at 2006 is 1.14%.
9
is immediately visible. Several areas stand out as especially dense, including the fertile areas in and around the Rift Valley Lakes in SNNP, and the eastern highlands of Hararge.
Figure 3: Population Dot Map showing total estimated population for 2004 based on 1994 census.
Source: CSA (2006), Atlas of the Ethiopian Rural Economy. Addis Ababa.
3.2. Population Growth and Urbanization Trend Table 1 shows population growth rates for the two most recent five year periods, presented at the national and regional levels. Growth rates are shown for rural and urban populations in addition to overall growth rates. For both periods, the growth rates for rural and urban population were not similar. The urban population growth rate was above 4%, while the growth rate for rural areas was around 2,6%. While growth rates are similar in many places, they are especially high in SNNP, for both rural and urban populations and for both periods.
Table 1: Urban and rural population growth in Ethiopia at national and regional levels, 1995-2005 Geographic Area 1995-200 2001-2005 Rural Urban Total Rural Urban Total National 2,74 4,38 2,92 2,57 4,10 2,73 Tigray 2,5 4,9 2,85 2,3 4.5 2,67 10
Afar 2,2 3,8 2,39 2,0 4,1 2,20 Amhara 2,6 4,9 2,86 2,5 4,4 2,67 Oromia 2,8 5,1 3,09 2,6 4,7 2,87 Somali 2,3 4,5 2,58 2,3 4,5 2,63 Benishangul Gumuz 2,5 4,7 2,57 2,4 4,4 2,54 SNNP 3,1 5,3 3,26 2,8 4,8 2,92 Gambela 2,2 4,7 2,62 2,2 4,4 2,57 Harari 2,5 4,2 3,50 2,6 3,8 3,49 Addis Ababa -- 2,9 2,90 -- 2,8 2,8 Dire Dewa 2,2 4,8 4,00 2,3 4,3 3,8 Source: CSA (2004), Projection based on 1994 Ethiopian Population and Housing Census. Addis Ababa. Different paths of population expansion can be compared with a population index, which is defined as a “normalized population size” with index value of 100 in the base year 1967. Figure 4 displays the growth paths of urban, rural and total population of Ethiopia as stated in terms of their index. It shows that urban population increased much faster than both the rural and total population during the last four decades. Since natural increases in urban population was not greater than for the rural population, the explosion of urban population shown on the top curve in the figure is due to the migration of people from rural to urban areas.
11
Source: CSA (1967/68-2007/8), Statistical Abstract. Series issues from 1967/68 -2007/2008. Addis Ababa.
3.3. Size of Urban Population The rate of urbanization depends on the definition of urban areas. The standard approach in measuring urbanization is the definition of settlement concentration, economic activity criteria and administrative classification (boundary of administrative unit) used by CSA. An urban centre includes, regardless of the number of inhabitants, “all capitals of regions, zones and weredas (districts), and localities with urban dwellers association (kabeles) whose inhabitants are primarily engaged in non-agricultural activities”; it also includes “settlements with more than 2000 inhabitants”. This approach has problems if the three criteria (administrative, nature of economic activity and settlement concentration) are used inconsistently over time. There can be over or under counting of townships.
The other option put forwarded is agglomeration index, which is based on three factors: population density, the population of a “large” urban center, and travel time to that large urban center. This approach assumes that “a city with a population size of 100,000 or more and a population density of less than 150 people per square kilometer does not exist”.
Using these two approaches the current urbanization rate is 16% as in the case of the first approach or 14,7% as in the case of agglomeration index. If we do not take the rates as absolute truth, rather as treating them as simply indicative of general broad trends, then the marginal difference between the urban percentage presented by the two approaches is not big. At the broad, aggregate national level, both rates can be considered as reasonable. The difference between the two approaches is important in understanding the spatial distribution of urban settlement (see section 4). In what follows I shall discuss the rate of urban population growth, how the rate has changed over time, where the population growth is taking place and comparison of levels of urbanization between countries.
From the viewpoint of the growth of urban population (Table 2), rapid urban growth can be divided into three periods. In the first period, 1960-1975, there was a considerably faster growth of urban population. The total population during 1960-75 period increased by 40 per cent as the period was relatively free from internal civil war and better agricultural growth. The period recorded a faster urban growth, the urban population rose by 113 percent. In the period 1975-1984, the percent variation in the total population growth was almost halved
12
down compared to the earlier period. This period was marked by natural calamities like hunger and severe internal civil war. The increase in urban population was 40 per cent compared to 113 per cent increase of the earlier period. Compared to 1984, there was modest increase in the urban population in 1994 recording 88,9 per cent variation, but this modest increase declined to 55 percent in 2008.
Table 2: Growth of Urban Population in Ethiopia, 1960-2008 Total Urban Urban Population Population Per cent population Per cent as per cent of total No of cities Per cent Year (mill) variation (mill) variation population and towns variation 1960 22,8 -- 1,5 -- 4 384 -- 1975 32,1 40,8 3,2 113,3 7 549 43,0 1984 40 24,6 4,5 40,6 11,4 710 29,3 1994 53,4 33,5 8,5 88,9 13,7 925 30,3 2008 79,2 48,3 13,2 55,3 16,7 927 0,2 Source: CSA survey of 1960 and 1975; CSA Census of 1984, 1994 and 2007
From the point of the growth of urban population, the year 1975 seems to divided the period 1960-2008 into two distinct phases. The per cent increase in urban population was high during the first phase, 1960-1975 compared to the period 1975-2008. The second phase commenced with the nationalization of urban and rural lands by the socialist military government and the kind of the rural biased policy continued during the agricultural-led development strategy of EPDRF government. Both governments have followed rural-bias policy compared to the urban bias policy of the monarch government of Haile Selassie.
3.4. Comparing Level of Urbanization between Ethiopia and Africa How fast has urbanization proceeded in Ethiopia? What is the level of its urbanization? A comparative analysis can provide some answers to these questions. Tables 3 and 4 shows the urbanization process in Ethiopia compared with Africa as a whole and some selected countries of Africa. Table 3 displays urban population as percentage of the corresponding total populations of the countries. The level of urbanization in Ethiopia is low by African standard. Table 3 compares the situation in Ethiopia, Western Africa (19 countries), Kenya and Ghana over the period 1960-2010 in this regard. The present level of urbanization is 17 % in Ethiopia, 51 % in Ghana and 40 % on average in Africa.
Table 3 Urban population percentage comparisons between Ethiopia, Africa, Western Africa, Kenya and Ghana, 1969-2010 Percentage urban (%) 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 13
Africa 18.7 21.3 23.6 25.7 27.9 29.9 32.0 34.1 35.9 37.9 39.9 Ethiopia 6.4 7.6 8.6 9.5 10.4 11.5 12.6 13.9 14.9 16.1 17.6 Kenya 7.4 8.6 10.3 12.9 15.7 17.0 18.2 19.0 19.7 20.7 22.2 Western Africa 15.2 18.7 21.4 24.2 27.3 30.2 33.2 36.0 38.8 41.7 44.6 Ghana 23.3 26.1 29.0 30.1 31.2 32.9 36.4 40.1 44.0 47.8 51.5 Source: Population Division of the Department of Economic and Social Affairs of the United Nations Secretariat, World Population Prospects: The 2006 Revision and World Urbanization Prospects: The 2007 Revision, http://esa.un.org/unup, April 2010.
Even if the level of urbanization in Ethiopia is low by African standard, the rates of urbanization in Ethiopia are exceptional (Table 4). Until 1985 the growth rate was more or less the same as the whole Africa. But in the last two decades the growth rate was very fast in Ethiopia, while the rate largely decline in other African countries (despite the rural bias policy). This can be for two main reasons: first, the fertility rate in Ethiopia was still high while it declined in other African countries; Secondly, increase in urban growth is rapid initially, when the level of urbanization starts from low level. “One percent of rural out- migration generates an additional six percent to the urban growth rate when the level of urbanization is around 14 %, and only one percent when the level of urbanization is 50 %. So, one might expect a higher urban growth rate in Ethiopia than in West Africa, which is much more urbanized”. This brings us to the issues of the relative contribution of natural increase (surplus of births over deaths) and rural–urban migration (surplus labor mobility).
Table 4: Urban growth rate comparison between Ethiopia, Africa, Western Africa, Kenya and Ghana, 1960-2010 Urban annual growth rate (%) 1960- 1965- 1975- 1980- 1985- 1990- 1995- 2000- 2005- 1965 1970 1980 1985 1990 1995 2000 2005 2010 Africa 5.10 4.65 4.43 4.30 4.16 3.87 3.52 3.38 3.31 Ethiopia 5.87 5.16 3.57 5.04 5.21 5.22 4.20 4.09 4.29 Kenya 6.32 6.96 7.65 5.34 4.92 3.90 3.44 3.60 3.99 Western Africa 6.46 5.26 5.22 4.99 4.80 4.48 4.31 4.01 3.77 Ghana 5.07 4.13 2.70 4.45 4.94 4.70 4.20 3.90 3.48 Source: Population Division of the Department of Economic and Social Affairs of the United Nations Secretariat, World Population Prospects: The 2006 Revision and World Urbanization Prospects: The 2007 Revision, http://esa.un.org/unup, April, 2010.
According to UN data, in the 1960s, the annual urban population growth rate was 5,5% (other sources claim 6.3 percent (Kloos and Aynalem, 1989). During this period the annual rate of rural-urban migration was 3.8 percent (Abate, 1989) and half of the population of the 14
towns with more than 20,000 population were found to be migrants in 1971 (Kloos and Aynalem, 1989).
The start of the modest decline in rate of urban growth since the early 1970s (3,57 according to UN data) continued at the same rate/level until 1980. This relatively low decline is mainly explained by 1975 land reform program. The nationalization and distribution of rural lands served an incentive for farmers to stay in the rural areas and there were also restrictions to live in urban areas (Baker 1996). The rate of voluntary rural-urban migration was lower in the mid - 1970s and 1980s. For instance, the urban growth rate of Addis Ababa was 3.0 percent (Mulugeta, 1985) and the rate of migration to Ethiopian towns was slightly above 1 percent (Abate, 1989).
Since 1985 the growth rate increased to 5,21%, the rural-urban influx related to mid 1980s famine and the consequent resettlement programs, and 1988-91 sever internal civil war and displacement of population led to growth in urban population. In the early 1990s the growth rate was 5,22%. With the coming of relative political stability in the second half of the 1990s, the growth rate decline to 4,2% and this rate is more or less maintained until the time of the research for different reasons of rural-urban migration and population movement.
15
Source: Population Division of the Department of Economic and Social Affairs of the United Nations Secretariat, World Population Prospects: The 2006 Revision and World Urbanization Prospects: The 2007 Revision, http://esa.un.org/unup, April, 2010
Table 5 shows that, while Ethiopia and West Africa have comparable total and urban population growth rates, the rural population grows two times faster in Ethiopia than in West Africa. Only one tenth of the natural growth of the rural population is absorbed by urban areas in Ethiopia, against two thirds in West Africa.
Table 5: Population growth comparison between Ethiopia and West Africa, 1960-2010 1960- 1965- 1970- 1975- 1980- 1985- 1990- 1995- 2000- 2005- 2010- 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 Total annual growth rate (%) Ethiopia 2.57 2.68 2.73 1.66 3.13 3.27 3.29 2.81 2.59 2.51 2.40 Western Africa 2.37 2.50 2.61 2.88 2.92 2.95 2.86 2.77 2.58 2.41 2.28 Rural annual growth rate (%) Ethiopia 2.32 2.46 2.53 1.44 2.90 3.00 3.00 2.58 2.32 2.16 1.95 Western Africa 1.54 1.81 1.88 2.07 2.09 2.09 2.01 1.85 1.62 1.38 1.19 Source: Population Division of the Department of Economic and Social Affairs of the United Nations Secretariat, World Population Prospects: The 2006 Revision and World Urbanization Prospects: The 2007 Revision, http://esa.un.org/unup, April, 2010.
4. Migration and Urbanization Pattern
4.1. Level and Forms of Migration
The driving force behind the rapid growth rate could not be the natural increase within the towns and cities themselves since considerable number of people have not yet started living in the urban areas. The main driving force behind the rapid urbanization in the country is the rural-urban migration, including reclassification.
Study on census data on places of birth and duration of residence (Figure 6) shows that rural-urban migration is the driving force of urbanization in Ethiopia. Migrants constituted almost half of the urban population for the past twenty years and the majority of them came directly from rural areas even if there was an increase in the urban-urban migration (Table 5).
16
Source: CSA Census of 1984, 1994 and CSA 1999 and 2005 Labour Force Survey
In 1984, out of the total population of 34, 5 million people enumerated by the census, 5,6 million persons (16,4%) were reported as migrants. The 1994 Census and the 1999 Labour Force Survey identified about 2,3 million to 2,5 million people who stated that they had lived in the places of current residence for no more than five years (Golini, A., Said, M. et.al (2001). Of this between 60-65% moved from rural areas. The structure of the major migration route shows that the rural-rural flow consists nearly a million people and the rural-urban flow over half a million people. The largest flow inside the rural area is not surprising since 85% of the population lives in the rural areas. There is also a significant flow (about 370,000 people) moving from urban to rural areas and again more than a million people moving from urban to urban areas. Female population mainly dominates the flows. About 55% of those who move are women and these form about 57% of those who leave from rural area (Golini, A., Said, M. et.al., 2001).
Rural-urban migration is widely considered as a rational behaviour and the decision to migrate is based on the analysis of costs and benefits. The Todaro model postulates that 17
migrants respond to rural-urban differences in expected rather than actual earnings and migration occurs until the expected rather than the actual incomes are equal across regions (Todaro and Smith, 2009). According to this general framework of choice weight is given to factors that affect rural income (land scarcity, government taxation) and factors that affect costs of migration (opportunity cost, cost of living and transport cost). If the cost of migration is high or if the expected income is low the decision to migrate also changes. This type of generalization may true in the case of the difficulty of getting access to farm lands as a result of growing number of young adult labour force in the rural area. Each year there are new entrants in the rural labour market as a result of the growth in the young adult labour force. Increase in the young adult labour is associated with decline in land for crop cultivation and this consequently leads to landless, underemployment or unemployment of the labour force. Migration to urban areas is an alternative. In case of regions affected by severe drought migration is not a choice but a necessity.
Table 6: Forms of Migration in 1984, 1994 and 1999 Forms of Migration 1984 Census 1994 Census 1999 Labour Force Survey Rural-Rural 55,8 48,9 37,6 Rural-urban 28,7 24,8 23,5 Urban-rural 2 7,3 15,7 Urban-urban 13,5 18,9 23,2 Source: CSA 1999 LFS analytical report
Studying the place of birth shows that all migration is not to towns. According to the 1984 census, some are rural-to rural (58%) and rural to urban (28%) and urban to urban (14%), and urban rural (2%). Rural-rural migration was twice of the rural-urban migration. As can be seen from Table 6 the largest form of migration is seen in the rural-rural category in all the three reference periods (1984, 1994 and 1999) comprising of 55,8, 48,9 and 37,6 percent of their respective total population. Rural-urban migration took the second position with a decreasing percentage share of 28,7, 24,8 and 23,5 percent in 1984, 1994 and 1999, respectively. Urban-urban and urban-rural forms of migration took the 3rd and 4th positions (CSA 1999 LFS analytical report).
Even if there was a decreasing trend in rural-rural migration (from 55,8 percent in 1984 to 37,6 percent in 1999), rural- rural migration was the dominant form of migration among most of the regions in Ethiopia. According to my framework of analysis, rural-rural migration 18
affects the urbanization process. “Movements between subsistence and plantation agriculture, between dry farming and irrigated areas, between villages subject to low correlations in their incidence of droughts, can play roles quite comparable to the strategies modeled with respect to rural-urban migrations.” It contributes to town formation and can increase or decrease their number (status of classification). Increase in the number of towns as a consequence of overgrown villages is the product intensive rural-rural migration.
The causes for rural-rural migration are related to the nature of demographic allocation of resources and stagnation of the subsistence economy. In my study area land could be acquired through inheritance, parental sharing, redistribution, forest clearing, renting and share cropping. These forms output increase through area extensification and the very nature of the land resource reallocation increases rural-rural migration. As the household multiplies and when land was no more accessed through demographic means, households are forced to intensify their production through intensification. Extensification and/or intensification (two non-market methods) could not increase production in the long run as there is a limit to area expansion and household labour productivity (Malmberg and Tegenu, 2010). In the absence of investment and change in technology, households will be forced to engage in the non-farm sector and this leads to an increase in the number of town markets.
4.2 Growth in Number of Towns and Size Distribution
An important feature of urbanization in Ethiopia is growth in the number of towns and shifts in the hierarchy of settlements. To get a comparative view of the pattern of growth of towns of different sizes I have classified towns on the basis of their population.
Earlier surveys particularly the three census classify cities and towns into a number of population size categories. To qualify as a town a place should have 2000 or more inhabitants or it must enjoy a statutory label of town administration. District government centre which do not satisfy the eligible criteria of 2000 can be considered as town if urban dwellers’ association are established in the locality.
Table 7: Town and city size distribution in Ethiopia, 1960-2008 Per cent variation, Year 1960-2008 Settlement Class and 1960- 1984- 1994- Hierarchy population Range 1960 1975 1984 1994 2008 1984 1994 2008 Super large cities I. 1000000+ 0 1 1 1 1 100,0 0,0 0,0 Large Cities II. 500000- 0 0 0 0 0 0,0 0,0 0,0 19
999999 Medium-sized III. 200000- cities 499999 1 0 0 0 3 -100,0 0,0 300,0 IV: 100000- Large towns 199999 0 0 0 3 8 0,0 300,0 166,7 V. 50000-99999 0 3 10 9 15 1000,0 -10,0 66,7 Medium-size VI. 20000-49999 5 8 13 30 61 160,0 130,8 103,3 towns VII. 10000-19999 5 22 38 69 145 660,0 81,6 110,1 VIII. 5000-9999 20 73 78 123 208 290,0 57,7 69,1 IX. 4999-2000 43 107 198 294 325 360,5 48,5 10,5 Small Towns X. Less than 2000 310 335 372 396 161 20,0 6,5 -59,3 All 384 549 710 925 927 84,9 30,3 0,2 Source: CSA survey of 1960 and 1975; CSA Census of 1984, 1994 and 2007
Table 7 illustrates the distribution of cities and towns classified by the number of population size categories. The towns are classified according to their population at the time of each census and by tracing the changes in each class regardless of the cities and towns included in it. A look at the total number of towns illustrate that, the year 1994 seems to divide the period of rapid urban growth into two distinct phases. The first phase, 1960-1994, recorded a larger increase in the total number of towns. In 1960 there were about 384 towns and this total has grown to 925 in 1994 (for the spatial distribution of towns see Figure 7). The total number of towns during 1960-94 increased by 90 per cent as compared to 2 per cent during the second phase, 1994-2008. The years 1994-2008 recorded the lowest per cent increase in urban number of towns. By the 2007 census only two new towns were considered as urban areas by the census decade preceding the census.
Why was there an increase in the number of towns during the period 1960-1994? Generally this period was characterized by modernization of the bureaucracy, revolution, nationalization of rural and urban lands, central planning, villagization program, internal civil war, famine and resettlement programs. It seems that increase in the number of towns was a consequence of overgrown villages and most growth occurred in the years 1960-84. From the data available it is very difficult to know how many of the towns that enter the data grow from settlements physically in existence for many years, prior to the time they pass the appropriate threshold of population, that is 2000 inhabitants. Entirely new towns also come into being, particularly those related to government settlement programs. It seems that the urban system is characterized by parallel growth (of unplanned and planned settlements).
20
Figure 7: Location of Towns and Roads in Ethiopia in 2004
Source: CSA (2006), Atlas of the Ethiopian Rural Economy. Addis Ababa
Population pressure and decline in agricultural productivity led people from northern, central and eastern highlands to move to uninhabited areas or cultivate the slopes at the edge of the lowlands (Abate 1992). In the period 1950-1974 it is conservatively estimated that the newly settled areas increased land under cultivation by about 25% and more than one million people were involved in the move (Wood 1982). Subsistence farmers dominate this type of migration.
In the period under discussion, in addition to the spontaneous redistribution of people, there were attempts to resettle people on the basis of government plan. In the late 1950s and early 1960s, the government planed to settle vast number of farm households on the state lands. The purpose was to relieve the population pressure in the over-crowded areas, and at the same time to increase government revenue through taxation (Rahmato 1989). In 21 schemes more than 6500 new farm units were created. The pace of resettlement has accelerated in the 1970s and a Settlement Authority was established in 1976 responsible for planned settlement. Many schemes were established in 1976 and 1978 and by the end of 1978,
21
the Authority had a total of 20435 households (75766 persons) in all its established schemes (Rahmato 1989). According to the official government report, there were 83 settlement schemes with a population of 40000 families or 160 persons by 1984 (Rahmato 1989) .
After the 1984/85 famine, a new phase of greatly increased resettlement program was started. The government decided to move about one and one half million people from the famine stricken areas of north, central and south-eastern parts of Ethiopia. By the end of 1987, a total of over 600000 people were located in the south-western provinces. This immense redistribution of population involved 1.6% of the country's rural population and it has inflated the population of the receiving areas (Rahmato 1989).
There is a dramatic decline and stagnation in the period 1994-2008, after the program of ethnic decentralization. In the period, 1994-2008, the total number increased by two percent compared to 43, 29, 30 per cent increase during 1975, 1984 and 1994 respectively. Since 1994 the number of towns hardly increased even if the period recorded an absolute increase in the percentage of urban population from 8,5% in 1994 to 13,2% in 2008.
The reason for lack of an increase in number of towns but an increase in the size of the already existing towns can be a result of changes in the economic structure of towns: growth in the non-farm economy (among rural towns of vii-x categories) and increasing distributive service functions of city and medium urban centers. It can also be related to administrative changes, merging of local towns and reclassification problems. Their negative growth since 1994 might show the limits that reached in the multiplication of subsistence households, shifts in labor force growth and absence of settlement programs by the state.
4.3. Shifts in Settlement Hierarchy: Growth in Number of Medium and Large Towns
Of the 927 number of towns only 15 have a population more than 50000 inhabitants, and 8 over 100000, and only 3 over 200000. The primate city Addis Ababa has a population of 3,1 million and is ten times larger than the second city of Dire Dewa (for list of cities over 20000 in the three census see Annex). Addis Ababa gathers only 4% of the Ethiopian population while the remaining 927 towns gather 12,7% of the population. There is a glaring absence of second rank cities and dozens of regional capitals of second class size (see Table 7) taking into account of the population of the country, being the second most populated country in sub- Saharan Africa. The absence of large and/or super large cities is one of the defining 22
characteristics of urbanization pattern in Ethiopia. The location of most of the towns in the Classes VII and X shows the predominance of the agriculture economy.
A look at the pattern of growth of towns of different sizes show that the period 1960-84 witnessed an increase in the population of class V to IX, with the largest per cent increases in class V. In 1984-1994 the largest increase was in class IV. In 1994-2008, one observes a shift of a large number of towns from lower classes to classes of III to VII. It seems that since 1994 there is faster growth rate in the middle class towns and negative growth particularly in class X. At the same time one observes an increase in the number of population of the already existing towns (see Annex). The second characteristic of the urbanization pattern is related to an increasing trend in the number and size of middle class towns (including non-farm based towns). The three census have not changed the definition of an urban area to affect the inter censuses comparability of urban population.
Growth in the size of the medium and large towns coincided with the labor force growth in the country. Since the middle of 1990s Ethiopia has started to experience an increase in working age population and an associated decline in the dependent age population. The labour force has grown rapidly as a result of moderate decline in the fertility rate.
Source: Source: Population Division, DESA, United Nations 23
Despite this changes, in 2008 towns with a population of less than 5000 still represent 52,4 % of the total town, while medium and large size towns constitute 46,3%. That means despite growth in the number of towns and shifts in the classes, agricultural towns are the dominant types, and this is the third characteristics of urbanization pattern in Ethiopia.
4.4. Regional Pattern of Urbanization (Urban Concentration)
Regional distribution of urban population and the pattern of urbanization is presented in Table 8.The level of urbanization varies regionally, ranging from nearly 100 per cent in Addis Ababa to 8,7 per cent in SNNPR. All other regions, with no exception, have experienced greater growth over the 1994 to 2007 census period.
Table 8: Population estimates and distribution of towns by region in Ethiopia No. Region Number of Population 2007 Percentage of Percentage of cities and (in thousands) urban population urban population towns 2007 2007 1994 census Urban Rural 1 Tigray 74 854 3595 19,2 14,9 2 Afar 28 132 1286 9,3 7,5 3 Amhara 208 2299 17325 11,7 8,5 4 Oromiya 375 3691 23613 13,5 10,6 5 Somale 69 768 3676 17,3 13,9 6 Benshangul-Gumuz 13 64 576 10 7,8 7 SNNPR 149 1338 13983 8,7 6,8 8 Gambella 7 49 204 19,4 16,7 9 Harari 1 127 76 62,6 58,1 10 Addis Ababa 1 3059 - 100 98,7 11 Dire Dawa 2 308 104 75 68,3 Total 927 12689 64438 16,5 13,7 Source: CSA 1994 and 2007 census
Development in transport, expansion of commerce and small manufacturing industries, re- organization of administrative set ups and decentralization have greatly affected the ranks of various towns. Road development and the consequent increase in population density along the road corridors created a kind of urban network among the isolated communities typical of the 1984 urban landscape (see Figure 10).
Figure 10: Travel time to a city for 2007
24
Source: Schmidt and Kedir (2009).
5. Economic Base and Classification of Urban Centres
When a high wave migration encounters low level towns it is only part of the migration wave that is transmitted across the towns. The exact behavior of attraction and repulsion 8 depends on the economic level of the towns and speed of the migration wave. One important property which explains attraction/repulsion is the economic function of the towns and their absorption capacity (i.e., impedance of the towns).
Classification of towns on the basis of their functions is problematic due to their complex functions: administration, defense, culture, production, communication and recreation. While a combination of these functions is performed by any give city or town, one function usually tended to overshadow the rest. Thus a concept of functional specialization is added to further
8 For a detail account on migration attraction and repulsion based on individual motivation framework data of 1994 census see http://www.irpps.cnr.it/etiopia/sito/progetto3.htm
25
explain the structure of urban functions (Smith, 1965). Towns with specialized functions can offer complex commercial services such as wholesaling, large-scale banking, and specialized retailing.
Specialization can be measured in two ways. As some quantitative studies do, specialization can be studied by looking at the occupational structure of the town’s labor force. Occupational data can reflect the kind of activities in which the town’s residents are engaged into: industrial, service, political, social, etc. activities. The question is how large a percentage of the labour force must be employed in a particular activity to attribute a specialized function to the town? There can be different threshold values for classification based on specialization: 30%, 40% or 50% of the employed labour force (Smith, 1965).
Another complementing criterion for specialization is the relationship between towns in the functionally specialized groups and their hinterlands. Some functions performed by the town are purely for the surrounding hinterlands (Christaller dubbed them as “central functions”, (Saey 2008). Towns performing central functions have non-tradable service sectors and they are not thus specialized. Towns which export their product out of the local area do not constitute a central place. They have tradable service sectors based on comparative and absolute advantages.
Finally, in classifying towns, it is important to see differences in the phases and levels of their development. Initially towns may develop as market centre using agricultural products. These towns can develop if there are conditions of labor force growth and road development. Infrastructural development at a later stage such us public buildings and interest in town land rent (by business people) leads to further growth of the towns. Establishing administration by itself does not increase in the number and economy of towns. Major routes of transport and communication lines not only stimulate economic growth but also accelerate the process of urbanization. City/town size distribution may show the level/phase of town development.
Without giving priority precedence to any of the classification criteria and concepts discussed above, the following grouping types of Ethiopian urban centers are identified: village towns along road sides (are villages less than 1000 but are classified as towns), non- tradable urban centers, tradable centers and cities.
26
5.1. Subsistence Towns
Subsistence towns are defined as those whose existence is based mainly on the services they furnish to the surrounding rural areas. The linkage of these towns to the rural side is not generated as a result of the technological change in agriculture as in the case of green revolution. The linkage is as a result of the coping strategies of subsistence households and their need for exchange of products.
As the marginal productivity of the subsistence labour decreases, due to continued increase in child-rich households and the consequent limitation in land-use changes (extensification and intensification), subsistence labour starts to depend on the market for the sale and purchase of production inputs and satisfaction of wants.9
Despite spatial differences in the population pressure, there appears to be different ways in subsistence households’ involvement in the exchange market. Initially when area expansion ceases to be the means of increasing production and when their land size holdings decline they start to produce high value crops using inputs mainly of labour. They, for instance, produce khat cash crops for sale. This contributes to the market integration of the different components of the economy. The market forces of cost and demand which transact produced item of khat are now functionally linked with those that transact the labour and land resources producing khat. A change in khat price "feeds back" on the rent price of khat land and the wage price of subsistence farm labour producing the khat.
The second alternative which contributes to market integration is subsistence households’ involvement in land market. When the return to intensive labour inputs becomes less (or when the start to receive low price for the cash crops), they start to expand production through land rent or share cropping. This is a kind of area expansion but through market mechanism and this leads to the development of land input market. Land becomes such a marketed commodity, whose ownership or use is something bought and sold for a money price determined by the same forces of market exchange which determine prices of labour and material items.
9 Subsistence labour/households are selling their products in the market even if they are primarily interested in use value. Not all of them are self-sufficient in everything they need. For instance, they sell grain to buy cooking oil and salt. This kind of exchange is about meeting needs, not making money as an end in itself.
27
When the return from rented land becomes low, subsistence households start to sell the household labour. A household head start to engage in agricultural wage employment (seasonal or causal) in order to maintain or compensate the level of food and cash needs of the household.
A study on links between market towns and villages in Ethiopia shows that “households undertake significant proportion of their economic transaction in local market towns. These localities are the site for about half the purchases of inputs used in agricultural production, from a quarter to three-quarters of sales of crops and livestock. They are the primary location of the sale of artisanal products, particularly by women. More than half of household purchases of consumables and various types of foods occur in these market towns. Strikingly, these are, largely, the only urban localities in which these rural households undertake economic activities.” (Dercon and Hoddinott, 2005).
The price of agricultural output and input are determined locally, within a range of eight to twenty kilometres distance. Agricultural crops and non-agricultural products produced by subsistence households (those produced for their own consumption and for sale in local market) are non-tradable, i.e., goods produced by all and everywhere. For instance, every one produces and sales similar product (grain and livestock trading) at the same season while there are few who can buy the products at the time. Every one produces and sales similar product (grain and livestock trading) at the same season while there are few who can buy the products at the time. Farmers therefore do not get the reasonable price for their product. Non- tradable services as distribution services that have not only higher transportation costs but also relatively few consumers of the service. According to Bradford and Kletzer (2005) “goods that are traded tend to be geographically concentrated (to capitalize on increasing returns to scale, access to inputs like natural resources, etc.), while goods that are not traded tend to be more ubiquitously distributed”.
Subsistence towns provide goods and services to their surrounding population. Perishable agricultural commodities and all non-food goods and services sold in the towns are also non- tradable, namely their output is constrained by local demand. Even if the supplies of non- tradable goods and services of the town are highly elastic, because of the low purchasing power of the subsistence households, the prices of the goods and services are fixed.
28
The prices of non-tradables are thus determined by supply and demand within the local economy of the towns. If one looks at their spatial distribution, agricultural towns illustrate lack of specialization in production structure. They are linked by road not through not as a result of specialization. One finds them randomly everywhere (they represent 52% of all the towns) depending on household multiplication in the given area 10 (Malmberg and Tegenu, 2010), and government villagization and/or resettlement programs. Their size varies according to the population pressure and availability of resources in their surrounding area.
5.2. Medium and Large Towns
In secondary urban centres other activities take precedence over agro-related activities, and agriculture in the surrounding countryside is based mainly on the demand for food created by the urban centre. Summary of the percentage distribution of Occupation of the Employed Population in urban areas of the country (aged ten years and above), shows that at the country level the highest proportion (24,8% ) of employed persons are engaged in Service workers, Shop and Market sales, and 24, 6% were engaged in elementary occupations. Next to this occupational group are persons engaged in Crafts and related trades, constituting 22,6% of the total employed persons. Skilled agricultural and Fishery, Technical and Associate Professionals, Clerks (“white collar occupations”) make up about 8,2%, and 5,5% and 4,8% percent of the country’s total employed persons, respectively. Plant, Machine Operators & Assemblers, Professional, and Legislator Senior Officials & Managers, make up insignificant proportion of the country’s urban total employed persons. They constitute altogether only 9,5% of the country’s urban employed persons.
10 The quality and quantity of resources available for a given household is determined by the number of household. For instance, in a situation where there is an increase number of household formation and number of children, the land size of enset and minor crops (sorghum and maize) are increasing to secure household food needs. This increase is done at the expense of cash crop and grazing lands (livestock) and forest areas. Household type (composition and size) and household numbers--both are the demographic factors—determine the size of the land for the cultivation activity and the proportional relationship between the crops. If, as a result of land scarcity, the area has no cash crop land, people ether sell the major and minor crops (food deficiency) or migrate to other regions to sell their labour.
29
Table 9: Employed population by major occupational groups 1999 and 2005
Total Employed Person
No. %
Year Legislator, Senior Officials Professional Technical and Associate Professionals Clerks Service workers, Shop and Market Skilled Agricultural and Fishery Crafts and Related trades Plant Machine , Operators & Assemblers Elementary Occupation N.S 1999 2702469 100 2,2 1,8 6,5 4,7 25,0 6,2 25,2 3,8 24,2 0,4
2005 3446092 100 2,2 3,6 5,5 4,8 24,8 8,2 22,6 3,7 24,6 0,2
Source: CSA National Labor Force Survey of 1999 and 2005
All of the functions performed by elementary occupation and all of the service sectors of the towns are purely for the surrounding regions and they have a non-tradable character. In other words, even if the urban centres are relatively large in size, the urban employment structure shows, the town economy is dominated by non-tradable services. That means their growth depends on government budget rather than on sale of services (labor productivity and specialization).
Most of the towns are public expenditure areas. The activities of some towns are very much affected by the socio-economic development projects commissioned by the government and donors. The government local expenditure budget has increased since decentralization of government power to district levels (Tegenu 2006). Regional level data on total public expenditure, disbursed investment credit, government subsidy program, donor budget support, etc. is need to discuss the contribution of public expenditure to rapid urbanization. The growth of the construction industry, particularly the recent growth of the residential sector, induced by the land grant and housing loan policies of the government, can contribute to the growth of urban settlements. Again local level data on the construction of condominium housing units, government offices and commercial projects is needed to assess the effects of the construction industry. If budget for employment is less compared to the growth in the labour surplus, there is unproductive employment in the sector.
There are as well goods which are tradable. Study on the livelihood zones in three regions of the country shows that part of the market sales constitutes tradable food cash crops (such as teff) and non-food cash crops (coffee, khat, flower horticulture). Part of the Crafts and related trades and most of the Plant, Machine Operators & Assemblers are related to mining and manufacturing activities of the urban centres.
Geographic concentration is interpreted as indicative of trade in that activity, in the sense that local production exceeds local demand in the given areas and the difference is traded 30
outside the area. This is clear in the case of manufacturing industries. Depending on their agro-ecological setting (such as rainfall amount, soil quality, topography and altitude) there is a concentration of tradable agricultural goods in these towns 11 and this can be a function of transportation facilities. Road development has been more rapid since the late 1990s and it facilitated the connection between trading centres, particularly in the agro-ecologically diversified regions. The liberal economy policy of EPDRF might have contributed to the growth in the tradable goods.
Table 7 on size distribution and the annex on dominant activities of towns show that tradable good centres (class category III-VI) are increasing both in number and size since 1994. Growth in size is not only related to transportation and government expenditure. It also coincided with growth in labour force. Since early 1990s Ethiopia has started to experience an increase in working age population and an associated decline in the dependent age population. The labour force has grown rapidly as a result of moderate decline in the fertility rate (Malmberg and Tegenu, 2010). Scarcity of farming lands and decline in agricultural productivity has led process of migration from rural areas to tradable urban centres.
5.3. Cities: Multifunctional in Nature
Metropolitan areas are the one or few largest population centres in almost any country, in which most industry, government and commerce are concentrated. A city is defined as settlements with more than 200,000 (for city hierarchy in terms of size of population see Table 7).
In 2008, Ethiopia had some 927 urban centers, of which only 15 towns have a population more than 50000 inhabitants, and 8 over 100000, and only 3 cities over 200000. The primate city Addis Ababa has a population of 3,1 million and is ten times larger than the second city of Dire Dewa. Addis Ababa gathers only 4% of the Ethiopian population while the remaining 927 towns gather 12,7% of the population. But this does that mean that Ethiopia suffers from an excessive development of its capital. The present growth rate of Addis Ababa, 4 %, is low
11 Agro-ecological factors are physical given affecting the nature of crops and cropping patterns. They affect what farmers can or should produce. In a higher altitude, for instance, farmers cannot produce teff crop, since the climate is suitable only for growing crops such as barely and wheat. In highland areas people can generally pursue an agricultural pattern of livelihood, while in the lowlands they can grow few crops and will be either pastoralists or agro-pastoralists, and so on. (For the regional patterns in tradable good services see Annex).
31
by world standards and Addis Ababa gathers only 4 % of the Ethiopian population, much less than Casablanca in Morocco, Johannesburg in South Africa, or Accra in Ghana.
Indeed, Ethiopia suffers from the relative weakness of its network of second rank cities, or regional capitals: one might expect that this country would count a dozen of such cities with 200 to 500 0000 inhabitants each, followed by some twenty cities (provincial capitals) with 100 to 200 000 inhabitants. These types of cities are multifunctional in nature (manufacturing, trade and transport, and service), manufacturing being as their predominant economic activity. The absence of medium and large size cities shows the lack of industrialization in all the years and existence of industrial policy (formulated since recently).
Given the glaring absence of medium and large cities, it would take more than five decades for the rural population pressure to stabilize if recent trends in surplus labor continue unabated (Malmberg and Tegenu, 2010). Already, the rural population pressure has reached a very high level incompatible with the availability of agricultural land. The way the peopling process of Ethiopia has been managed through government rural-bias policy during the last four decades is unsustainable over the long term.
6. Consequences of Migration-led Urbanization and Government Urbanization Policy
Even if urbanization is low in Ethiopia, it is growing rapidly by African standard and this has its own effects. A number of studies have identified the following consequences:
• It created unemployment, underemployment and lower labor productivity both in the formal and informal sectors. • Massive un- and underemployment, which is associated with low household income, causing widespread urban poverty. • It created demand for more services: housing, education, health, sanitation, transport, communication, energy and business services. • Because governments have less revenue to spend on the basic upkeep of cities and the provision of services, city and towns have become areas of massive sprawl- scattered development that increases traffic, saps local resources and destroys open space. Urban sprawl has become a specific feature associated with rapid urban growth in Ethiopia.
32
Despite difference in measurement criteria and scale of interpretation, studies on urban poverty found out that 47%-50% of the urban population is below the per capita poverty line. Urban households face particularly high costs related to food, h ousing, transport, basic services and health (on the level and different aspects of urban poverty including income, access to basic services and measures of exclusion see Kedir 2005, Gebremedhin 2006, Bigsten and Shimeles, 2008, Mulugeta, ed. (2008).
What is missing and not emphasized in the urban poverty studies is the wider context of the problem, namely the out and in migration of surplus labor. The causes of the surplus labor are related to the demographic/economic situations in the rural areas and economic impedance of towns (their absorption capacity). In the rural area surplus labor was created mainly through two mechanisms. The first is the situation of underemployment of the substance economy where household’s heads engage in non-farm activity as a result of productivity stagnation and marginal returns to household labor. The second mechanism is the growth new entrants in labor force and the difficulties of finding self or wage agricultural employment due to decline in man/land ratio. The inability of the agricultural sector to provide job to the surplus rural labor (see Figure 11) led to out-migration to urban centers. My findings seem to suggest that it is not expected income that motivates individuals or households to migrate to urban centers. Because individuals could not live on the minimum subsistence level, migration has become a necessity not a choice.
The Todaro migration model states that migrants respond to urban rural differences in expected rather than actual earnings (Todaro and Smith 2009). Relatively good income alone may not motivate households to move to towns if there is an improvement in the subsistence level. Rural development programs improving living conditions could discourage movement to towns. What I have observed in the case of Ethiopia is that the reforms, for various reasons, could not cope up with the growth in the surplus labor. Agricultural reforms, in whatever form, could not cope up with the growth of the labor force and the pressure of high food consumption households. So we have a situation of massive migration between rural-rural and rural-urban areas.
33
Source: Authors own calculation from CSA labor survey data and agricultural reports
The urban-rural migration is low compared to rural-rural migration for two major reasons. The first relates to the low economic level of the towns (compared to even other African countries) and their capacity to absorb the migration wave. There are three economic bases of the towns, identified on the basis of their functional nature. The first is the off-shot of the subsistence sector: subsistence households depend on the local market to complement their income through undertaking of non-farm economy. This has led to the transformation of settlements into small towns particularly along the roads. But the market is fragmented; there are lots of subsistence households selling products. The second is the tradable agricultural products which have become the economic base of medium and large size towns. Because of the technological stagnation of the agricultural sector and lack of transportation facilities the supply of tradable goods, particularly the food cash crops could not increase in volume as with demand. The third economic source of the town is the government budget and this fluctuates depending on expenditure priorities, sources and levels of government revenue.
The economic impedance nature of the urban centers is thus one reason for low level of migration into urban areas. Another reason why towns and cities could not reach to substantial
34
size, despite the growth in surplus rural labor and increasing of out-migration, can be rural biased policy of the Ethiopian government. Since 1974 the Military Government pursued rural bias policy consisting different programs: nationalization of urban and rural lands, distribution of rural lands to households, formation of farmer production co-operatives, villagization, resettlements, etc. Since 1991, the EPDRF Government pursued agricultural-led development strategy consisting different rural development program packages. The rural biased policy tied the mass of redundant labor to the land and since the demographic/ fertility behavior of the labor force did not change government rural reforms provided fuel to the momentum of population reproduction (Malmberg and Tegenu, 2007).
It was very recently, in 2004, that The Ethiopian Government formulated the National Urban Development Policy (NUDP), ten years after the problems of spontaneous and rapid urbanization reach the climax. The NUDP has two principal packages: Urban Development Package (UDP) consisting housing development program, MSE development program, land and infrastructure development program and urban-rural linkages. The Urban Good Governance Package (UGGP), the second principal package, consists of sub-programs that reinforce UDP.
The NUDP represents a comprehensive set of measures to promote urban development by providing equal opportunity and affordable services. A review of the policy ( Zerihun, 2008 ) identified the following gaps: the policy missed out the issue of financing the programs; does not stipulate appropriate time frame for attaining results; provision of road, water and telephone infrastructure are missing elements; accelerated housing program is not emphasized given enormous backlog; building local democratic institutions is not emphasized; the availability of adequate revenue for the urban administration is not thoroughly discussed.
The NUDP has clearly identified the consequences of rapid urbanization in Ethiopia and designed a program and institution to address the problems. Even if the gap mentioned by the review can be amended, it is doubtful that the policy can generate conditions for self- generated urbanization in Ethiopia. The policy finds solutions to the effects of rapid urbanization in Ethiopia (symptoms); it does not address the problem related to the root causes of migration: high rural fertility, increasing food consumption requirements, resource scarcity (particularly land), growth in new entrants in labor force, and lower economic capacity of the urban formal and informal sector to absorb surplus rural labor. Given the
35
demo-economic situation in rural areas and the economic impedance nature of the urban centers, how can a rural surplus labor be absorbed and agglomeration economies created?
7. Alternative Policy Strategy: Green Revolution and Industrial Decentralization
The premise of this study is that the problems of surplus labor can be solved through labor productivity, encouraging of specialization both in agriculture and non-farm economy based on comparative advantages, and construction of transport facilities. T he main objective is creation of employment for the unemployed and under employed (through input growth and final demand) and raising labour productivity (for the employed). The faster the surplus labour grows, the more pressure there is on job creation. To this effect, there is a need for parallel transformation of the subsistence and non-farm based economy in the rural areas.
The strategy of giving primacy to agriculture, on the ground that the majority of the people live in rural areas is a mere simplification of the reality based on ideological orientation. This approach does not work in conditions where people are migrating from one saturated area to another saturated area. Development should occur simultaneously in both rural and urban areas to ensure economic integration and reinforcing expansions for the purpose of absorbing surplus labour. One cannot solve rural problem without the urban and vice versa. Indeed this requires enormous resources and it may be easier said than done. But there is no other option which can help to catch up the race.
By looking at the nature, size, speed and need of the rural surplus labour; the uneven spatial distribution and unbalanced growth of urban centres; multiplication of subsistence households with higher food consumption requirements; and, reviewing the capacity of current government policy to counter the negative effects of the pressure drivers, I suggested green revolution and industrial decentralization as policy options to keep up with the race. Green revolution means introducing agriculture technology in rural areas: high-yielding varieties of grain, use of pesticides, and construction of irrigation physical structure facilities and management for harvesting two or three times per year and to avoid dependency on rain- fed agriculture. Industrial decentralization is defined as the spread manufacturing employment and enterprise management from major cities to rural towns through a phased transition.
State-led Green Revolution
36
In the case of subsistence economy, it may be necessary to design a green revolution program which gradually gets rid-off small holder agriculture in Ethiopia (for a short review of the Ethiopian government agriculture-led growth strategy, see, Dercon, Hill, and Zeitlin, 2009). The empirical results show that the subsistence sector is a deficit and unproductive sector (Tegenu, 2009). There is no need to sustain it through reforms aimed at developing its tillage/cropping systems and efficiency in a way that can be suitable for the soil and climatic conditions. The subsistence sector has to disappear or dissolve into the employment sector and this can happen mainly through the green revolution undertaken in Asian countries. Green revolution leads to a significant increase in agricultural productivity resulting from the introduction of high-yield varieties of grains, the use of pesticides, and improved management techniques. Green revolution requires government investment in infrastructure such as irrigation, land consolidation programme (and/or formalizing land market), and modern input provision with all the infrastructure and packet. To get quick positive effect of this policy measure it is necessary that the package is accompanied by family programming (demography and health programs, see WB 2007). It is only agricultural productivity of green revolution which has the capacity to feed the growing urban population in Ethiopia. Currently the urban population is 12 million, and it increases by a growth rate of 4,7%
Specialization instead of Diversification Concerning the surplus labour various literatures recommend the promotion of non-farm economy (WB, 2009) through provision of credit (micro finance), business support services in training and technical assistance, and provision of infrastructure such as electricity and water. As my study on the rural employment structure shows, the non-farm economy is dominated by non-tradable services, i.e., goods produced by all and everywhere. In the case of small trading rural towns, for instance, every one produces and sales similar product (grain and livestock trading) at the same season while there are few who can buy the products at the time. Even if the service sector is less land intensive compared to agriculture, it cannot absorb surplus labor if it is non-tradable. In the case of rural Ethiopia, its growth depends on government budget rather than on sale of services (labor productivity and specialization). If capital for employment is less compared to the growth in the labour surplus, there is unproductive employment in the sector. What is need is a shift to new occupations involving higher levels of skills and better technology. In other words, increasing labour surplus is
37
absorbed by transforming the structure of the non-farm employment, not by supporting continued diversification.
Industrial Decentralization This brings us to the issue of the need for industrialization, the transforming of the economic structure into higher level of productivity and employment creation. Economic literature underlines the importance of industrialization. The problem is adopting a standard industrial policy menu to the unique conditions of Ethiopia. Considering the size and growth of the surplus labour, growth in the number of small and medium towns, and the need for incentive to competitive production, I suggest fragmented/segmented systems of manufacturing production relocated in medium and small towns . Industralization is conceived here as the spread of manufacturing employment, shifting of future industrial growth from cities to secondary urban centres. Manufacturing industries include chemical, engineering, electronics, textile, metalworking, construction, energy, plastic, transport and telecommunication, and food and beverage industries. Manufacturing industries not only have the capacity to absorb the growing rural and urban surplus labor of the country, they have as well a strategic importance in saving foreign exchange and reducing dependency.
The manufacturing enterprise may be privately and/or collectively owned, jointly financed by the state and collective units, or wholly owned by the state but under local management. A planned shifting of manufacturing units away from cities and large size towns to medium and small size towns requires a decentralized organization structure. Currently the decision is central and top-down (Timan Altenburg 2009 and Kenichi Ohno 2009). The Ethiopian Industrial Development Strategy is not yet fully translated into policies, sector programs and action plans (for a review see Zerihun 2008).
Industrial decentralization provides an incentive to competitive production. Growth cannot be sustained for long without structural competitiveness. Sustainable growth is the outcome of concentration of economic activities in urban centres (following the model of economic geography), development of human capital (following endogenous growth theory) and innovative entrepreneurship (neo-Schumpeterian economic theory). Industrial decentralization is about the skill development of the young labour force and the process of economic driven urbanization. It introduces a systematic competitiveness in the economy as privately and/or collectively owned local and national manufacturing firms compete for production and export.
38
China is a good example in this particular case. Systematic competitiveness has the advantage to create structural stability since it provides sustainable economic development that has a capacity to manage social change without resorting to violent conflicts.
Phases of Industrial Decentralization Decentralization has two phases. In the first phase the federal and regional governments have to establish manufacturing industries in large and medium size towns. Currently there are 84 towns with a population of 20,000-200,000, and the current migration trend is towards these towns. In these urban centres the federal/and or regional government has to lay the foundation of large-scale chemical and agricultural manufacturing industries, instead of focusing on medium and small scale enterprises (MSEs). In the second phase of decentralization, the focus should be on developing of small scale enterprise in small towns (such as establishment of oil mills, soap manufactures and textile factories). There are 353 small towns with a population 5000-19999, providing mostly non-tradable services to the hinterlands. Development through decentralization and market liberalization involves the interplay of capabilities, incentive and institutions (for recommendation on the enabling reforms see Tegenu 2006).
39
Reference
Abate, A. (1989). “Internal migration and urbanization in Ethiopia.” Paper presented at the conference on Population Issues in Ethiopia’s National Development, Addis Ababa. Abate, A. (1992). Urbanization in Ethiopia: Past Trends and Future Prospect. Abstract, 1992 Altenburg, T. (2009), “Industrial Policy in Ethiopia”, Workshop, Industrial Policy in Developing countries”, DIE (German Development Institute). Bonn. Baker, J. (1996). Small Urban Centres and their Roles in Rural Structuring in Zegeye, Abebe and Pausewang, eds. Ethiopia in Change: London. Bigsten, A. and Shimeles, A., 2008, The persistence of urban poverty in Ethiopia: A tale of two measurements. Department of Economics. Göteborg University. Bradford, J. and Kletzer, L.G., (2005), Tradable Services: Understanding the Scope and Impact of Services Offshoring, in Brookings Trade Forum, 75-116. CSA (1967/68-2007/8), Statistical Abstract. Series issues from 1967/68 -2007/2008. Addis Ababa. CSA (Central Statistical Agency), 2006, Atlas of the Ethiopian Rural Economy. Addis Ababa CSA (2007). Statistical Abstract. Addis Ababa. CSA Survey of 1960 and 1975. CSA 1999 LFS analytical report CSA Census of 1984, 1994 and 2007 CSA Census of 1984, 1994 and 2007 CSA Report on the 2005 National Labour Force Survey CSA survey of 1960 and 1975; Deng , X., Huang,J., Rozelle, S., Uchida, E., (2008) Growth, population and industrialization, and urban land expansion of China, in Journal of Urban Economics 63 (2008) 96–115 Dercon, S. and Hoddinott, J, 2005, Livelihoods, Growth, and Links to Market Towns in 15 Ethiopian Villages. FCND Discussion Paper No. 194 Dercon, S., Hill, R.V. and Zeitlin, A. (2009), In search for a strategy: Revisiting Agricultural-led Growth in Ethiopia. University of Oxford. Gamst, F.C. (1970), Peasantries and Elites without Urbanism: the Civilization of Ethiopia, in Comparative Studies in Society and History, 12. Cambridge University Press. Gebremedhin, T. A. (2006) The Analysis of Urban Poverty in Ethiopia. The University of Sydney;
40
Golini, A., Said, M., Casacchia, O., Reynaud, C., Basso, S., Cassata, L., and Crisci, M., (2001) Migration and Urbanization in Ethiopia, with special reference to Addis Ababa. La Sapienza, Roma: Department of Demographic Sciences, University of Roma, and Addis Ababa: Central Statistical Authority. Henderson, J V, 1974. "The Sizes and Types of Cities," in American Economic Review, American Economic Association, vol. 64(4). Kedir, A. M. 2005, Understanding Urban Chronic Poverty: Crossing the Qualitative and Quantitative Divide. University of Leicester, Chronic Poverty Research Centre Working Paper 53; Kloos, H. and Adugna, A. (1989), The Ethiopian Population: Growth and Distribution. The Geographical Journal: V. 155, No. 1. March 1989 Ledent, J. (1982), The Factors of Urban Population Growth: Net Immigration versus Natural Increase, in International Regional Science Review. 7; 99. Malmberg, B. (2008), Demography and the Development Potential of sub-Saharan Africa. Current Africa Issues No 38, Nordic Africa Institute. Uppsala. Malmberg, B. and Tegenu, T. 2007, Population Pressure and Dynamics of Household Livelihoods in an Ethiopian Village—An Elaboration of the Boserup-Chayanovian Framework. Mulugeta, E. ed., (2008), Urban Poverty in Ethiopia: The Economic and Social Adaptations of Women. Addis Ababa University Press. Addis Ababa Mulugeta, S., 1985. Meeting the Housing Shortage in Addis Ababa: The Case of Housing Cooperatives. Master’s thesis, Addis Ababa University. NOP, Ministry of Economic Development and Cooperation, 2000. The Ethiopian Population Profile: 1999. Addis Ababa Ohno, K. (2009), Ethiopia: Political Regime and Development Policies Oluwasola, O. (2007), Social Systems, Institutions And Structures: Urbanization, Poverty and Changing Quality Of Life, Paper Presented At The Training Session Of The Foundation For Environmental Development And Education In Nigeria. ILE-IFE. Population Division of the Department of Economic and Social Affairs of the United Nations Secretariat, World Population Prospects: The 2006 Revision and World Urbanization Prospects: The 2007 Revision, http://esa.un.org/unup, April 2010 Rahmato, D. 1989, 'Rural women in Ethiopia: Problems and prospects', mimeo, University of Addis Ababa. 41
Saey, P. (2008) The Study of Cities: Historical and Structural Approaches. Retrived from http://www.lboro.ac.uk/gawc/rb/rb276.html ). Schmidt, E. and Kedir, M. (2009), Urbanization and Spatial Connectivity in Ethiopia:Urban Growth Analysis Using GIS, Ethiopia Strategy Support Program 2 (ESSP2), Discussion Paper 003. Silberfein, M. ed., (1998), Rural Settlement Structure and African Development. Oxford. Smith, R.H.T. (1965), Method and Purpose in Functional Town Classification, in Annals of the Association of American Geographers, Vol. 55, No. 3. Tegenu , T. (1996), The Evolution of Ethiopian Absolutism. The genesis and the making of the fiscal military state, 1696-1913. Uppsala Tegenu, T. (2004), “Socio-economic and Environmental Effects of Age Transition in Ethiopia: 1950-2000”, in Proceedings of the First International Conference on the Ethiopian Economy. Edited by, Ethiopian Economic Association, V.III. Tegenu, T. (2006), Evaluation of the Operation and Performance of Ethnic Decentralization System in Ethiopia. The Case of the Gurage People, 1992-2000. Addis Ababa University Press. Addis Ababa.). Tegenu, T. (2008), The Onset of Demographic Dividend Period and the Effects of Labor Force Growth on Ethiopian Rural Economy: A Study of Growth Policy Options. Paper presented at Expert Group Meeting (EGM) on Mainstreaming Age Structural Transitions (ASTs) into Economic Development Policy and Planning. Vienna. Tegenu, T. (2009), Explaining Drivers and Consequences of Population Pressure in Rural Ethiopia. Department of Human Geography, Stockholm University Teller, C., Gebreselassie, T., and Hailemariam, A. (2007), The lagging demographic and health transitions in rural Ethiopia: Socio-economic, agro-ecological and health service factors effecting fertility, mortality and nutrition trends. Paper presented at Session 104, Population growth and poverty linkages in Africa, Fifth African Population Conference, Union of African Population Studies (UAPS), 10-14 December, 2007. Arusha. Todaro. M.P. and Smith, S.C., (2009), Economic Development (tenth edition). London. WB (World Bank) 2007, Ethiopia Capturing the Demographic Bonus in Ethiopia: Gender, Development, and Demographic Actions. Report No. 36434-ET. WB, 2009, Ethiopia Diversifying the Rural Economy An Assessment of the Investment Climate for Small and Informal Enterprises October 6, 2009
42
WDR (World Development Report) 2009, Spatial Disparities and Development Policy, Outline, retrieved from http://siteresources.worldbank.org/INTWDR2009/Resources/Outline.pdf Wood, A.P. 1982 Wood, Adrian P. 1982, Spontaneous Agricultural Resettlement in Ethiopia 1950- 1974. In J. Clarks and L. Konsinski (eds.), Redistribution of Population in Africa. London. Zerihun, A. (2008), “Industralisation Policy and Industrial Development Strategy of Ethiopia”, in Taye Assefa, ed., Digest of Ethiopia’s National Policies, Strategies and Programs. Addis Ababa.
43
ANNEX : Towns of over 20,000 population during the last 3 census periods and their dominant activity
Region/Zo Dominant Economic ne Name Activity measured by employment share (in City Name 1984 1994 2007 thousands) Addis Ababa Central 1,423,182 2,084,588 2,738,248 South Dire Dawa Eastern 99,98 173,188 232,854 Trade and livestock Central CRT 78, EC 58, SWSMS Adama Special 77,256 127,842 222,035 56, SAF 27, livestock North CRT 30, SWSMS 21, EC western 14, TAP 5, trade in Gondar 80,675 112,249 206,987 consumer goods Southern EC 17, SWSMS 16, CRT Awasa 36,367 69,169 158,273 15, TAP 7, C 4, Khat North CRT 21, SWSMS 17, EC western 8, TAP 2, manufacturing, Bahir Dar 54,773 96,14 155,355 trade and Livestock South SWSMS 16, EO 9, CRT 6, Jijiga eastern 24,716 58,36 125,584 SAF 6, Khat Northern SSMS 19, ERT 17, EO 16, Mekele 62,668 96,398 122,85 C 4 South SSMS 25, EO 20, CRT 17, western TAP 5, C 3, Coffee and Jimma Special 60,218 88,867 120,6 khat North CRT 20, EO 15, SSMS 16, Dese easern 71,565 97,314 120,029 TAP 4, C3, Khat Southern 78 CRT, EO 58, SWSMS 56, 27 SAF, TAP 7, C 5, Shashemene 31,884 52,08 102,062 Khat, maize and teff Central 78 CRT, EO 58, SWSMS 56, 27 SAF, TAP 7, C 5, Bishoftu 55,657 73,372 100,114 Teff and flower South SWSMS 16, EO 9, CRT 6, eastern SAF 6, livestock and Harer Ketema 63,07 76,378 99,321 marbel Southern SWSMS 9, EO 6, CRT 5, Dila 22,864 33,734 67049 Khat and coffee South CRT 15, SWSMS 12, EO Nekemte western 28,703 47,258 76,817 11, TAP 4, Coffee Southern SWSMS 8, EO 7, CRT 4, Sodo 24,278 36,287 76,780 SA 3, Khat and livestock Arba Minch Southern 20,28 40,02 74,843 Fish and fruit Southern EO 10, SWSMS 7, SA 6, Hosaena - 31,701 69,957 CRT 6, Asela Central 32,954 47,391 67,25 wheat Debrebrehan Central 25,637 38,717 65,214 Livestock 44
North Debere Markos western 41,138 49,297 62,469 Teff North Kombolcha eastern - 39,466 58,642 manufacturing Adi Girat Northern 37,417 57,572 Sebeta Central - - 56,131 Livestock and flower North Debretabor western - - 55,157 Livestock Ambo Central - 27,636 50,267 Livestock and flower Burayu Central - - 48,864 Flower Arsi Negele Central - 23,512 48,092 Livestock South Robe eastern - 21,516 47,296 Wheat and Livestock Shire Enida Northern Silase 46,382 North Woldiya eastern - - 46,126 Akisum Northern - 27,148 44,629 Zeway Central - 20,056 43,61 Livestock and flower South Livestock and trade in Gode eastern 40,585 43,134 goods Adwa Northern - 24,519 40,502 South Gambela western - - 38,994 Weliso Central - 25,491 37,867 teff, wheat, flower South Negele eastern - 23,997 36,699 Livestock Meki Central - 20,46 36,214 Livestock, teff and flower South Chiro eastern - - 33,643 khat Butajira Central - 20,509 33,393 khat Alamata Northern - 26,179 33,198 South Goba eastern 23,052 28,358 32,916 wheat and livestock South Haromaya eastern - - 31,686 khat and livestock Yirgalem Southern - 24,183 31,468 coffee South Gimbi western - 20,462 31,311 coffee and marbel Wekero Northern - - 30,208 South Degehabur eastern 25,464 29,956 Livestock South Metu western - - 29,627 Coffee Mojo Central - 21,997 29,54 Teff South Dembi Dolo western - - 29,249 Gold Welikite Central - 28,856 Khat, Livestock and flower Areka Southern - - 28,293 Teff 45
Boditi Southern - - 27,684 Livestock FicheTown Central 21,187 27,487 Livestock Kulito Southern - - 26,85 Khat Hagere Mariyam Southern - - 26,6 Coffee South Livestock and trade in Kebridehar eastern 21,521 26,479 goods South Livestock and trade in Dolo Odo eastern - - 26,296 goods North Mota western - - 26,173 Teff South Agaro western - - 25,719 Coffee North Finote Selam western - - 25,388 Teff North Kobo eastern - 20,788 24,861 North Dangila western - - 24,564 Teff South Tepi western - - 24,489 Coffee Durame Southern - - 24,454 Coffee and Livestock Holeta Central - - 24,03 Teff, Livestock and flower South Shakiso eastern - - 23,924 Gold Mizan Aman South Kifele Ketema western 23,562 Coffee Maychew Northern - - 23,484 Gerbe Guracha Central - - 23,481 Teff and marbel Sawula Southern - - 23,37 khat and livestock North Chagini western - - 23,225 Teff South AdolaTown eastern - 22,937 Gold and livestock Southe Assosa western - 22,725 marbel and flower Sekota Northern - - 22,342 public expenditure Aleta wendo Southern - - 22,09 Gold and khat South Dodola eastern - - 21,504 wheat and livestock Genet Hara Central - - 21,504 North Humera western - - 21,387 oil seed North Wereta western - - 21,217 Teff North Injebara western - - 21,058 Teff Bonga Southern - - 20,855 Coffee North Debarq western - - 20,83 Teff
46
Asassa Central - - 20,647 Wheat Jinka Southern - - 20,522 Coffee and livestock North Bure western - - 20,406 marbel Total 21 43 88 Source: CSA 1984, 1994 and 2007 census and interview material Notes: P=Professional; LSOM=Legislator Senior Officials & Managers; TAP=Technical and Associate Professionals; C=Clerks; SWSMS=Service workers, Shop and Market sales; SAF=Skilled agricultural and Fishery; CRT=Crafts and related trades; PMOA=Plant, Machine Operators & Assemblers; EO=Elementary Occupation
47